Angle

A prospect sends 115 security questions, and the team hesitates between rebuilding everything and copying the previous client's questionnaire. Both fail for the same reason: they take the document for the knowledge. The first throws away what past research established because it exists only in a file that was sent; the second reuses the file believing it is reusing what it knew. A company only keeps what it knows in the form of facts that carry their origin, their date and their limits, consulted with the team that answers for them or copied with a note of where they came from — and it is that form, not the number of archived answers, that turns each new questionnaire into a contribution rather than an investigation.

Synthesis

The starting point is a cost experienced as inevitable: every client sends its own format, its own numbering, its own wording, and the research starts over even though the topics keep coming back. The first reading is that the work gets lost; the second, more uncomfortable, is that the reflex meant to save it — copying — loses it too, because an old answer has the shape of knowledge without its guarantees.

The same mechanism plays out between teams, and that is what gives the whole its reach beyond questionnaires: a rule copied into your own space stays frozen while the original evolves, without a sound. Consulting and copying are not opposed as the good and the bad move; each has its debt, and each shifts responsibility differently. Correcting someone else's knowledge follows the same logic: the correction is proposed to the owner.

The decisive change is a change of axis: filing by topic rather than by client. It brings together questions numbered differently, it exposes gaps instead of hiding them, and it lets you separate the evidence, which stays internal, from the wording, which goes to the client. The measured gain — a questionnaire 80-90% prepared in ten minutes — is not a gain in certainty: it moves human review onto the commitments.

What the whole says that no single note says: the loss of knowledge is not a failure of memory but a failure of container. The document is necessary for the client; it cannot also be the place where the company learns.

Tensions / contradictions

The adapted copy is defended elsewhere as the right way to pass between two domains, precisely because it avoids depending on the source; here, any copy is suspect as long as it doesn't say where it came from. Both hold if the provenance is written down, but one values the independence the other treats as a risk.

A second tension: consulting the rule with its owner assumes being told when it changes, and that signal depends on a team that doesn't necessarily know who relies on it. The reference's debt is paid by someone other than the one who took it on.

Questions

• How does a company spot the topics that come back often enough to deserve a verified fact, before having answered several times?
• Who, in an organization, bears the cost of verifying dated facts again when no questionnaire demands it?
• Does the same filing work for knowledge that can't be reduced to facts — know-how, a commercial trade-off?